An Analysis Of The Word Order Pattern In The SI Target Language And Its Underlying Reasons In The Language Combination Of English And Chinese

As simultaneous interpretation (SI) is real-time translation of sense, researchers in interpreting studies have taken note of the topic of how interpreters cope with word order differences of the source and target languages (SL and TL). The focus of this research topic in the West has shifted from description to exploration of strategies adopted by the interpreters and the underlying reasons, and the language combinations concerned have been predominantly Indo-European. Chinese researchers, on the other hand, have also demonstrated great interest in the topic as evidenced by the citations of Chinese literature on SI. Unfortunately, those most cited literature, owing to limitations on length among many factors, offered advices on the way to tackle word order differences in Chinese-English SI, or evidence-based descriptions of the increase of interpreters'ear-voice span (EVS), while no sign of the shift in the West has made itself manifest. With that in mind, this dissertation sets out to explore the word order patterns of segments of noun phrases constructed by head words and modifiers in the target discourse of professional interpreters, and based on uniform information density (UID) and the relation between entropy and information profile, to investigate the underlying reasons behind such patterns in an attempt to construct flow chart models of the mental process of interpreters'choices.This dissertation is based on controlled empirical study with professional subjects interpreting simultaneously in both directions speeches that have such variables as speakers'topic, accent and speed under control in standard booths. Quantatative and qualitative analyses are conducted following the transcription of audio files and the calibration of time. Measurement of EVS of the concerned segments is based on judgment of their correctness and TL word order patterns. After calculation of the mean time lags of the two TL word order patterns, the uncertainty of and the memory load induced by the segments are analyzed. Within the framework of UID and the relation between entropy and information profile, a set of formulas for the two basic structures SV and SVO is derived, as well as a set of time rules for the formulas. Conclusions are drawn on the basis of TL discourse analysis that when the information profile of SL has been decided, most interpreters choose a word order pattern that prioritizes the interpretation of the decided profile in TL, i.e. keeping the TL word order consistent with that of SL. But when the SL information profile has not been decided, interpreters predominantly prioritize waiting to decide the SL profile. Though interpreters mostly give top priority to whether the SL information profile is decided in choosing the TL word order patterns, a few sometimes rests their word order decision on the memory load induced by the segments, giving up on this information-profile basis. Memory load is a potential factor influencing the interpreters'word order patterns. Most subjects'EVS performance is in harmony with the time rules of the formulas in SI from English into Chinese, but in the other direction, meaning has a more significant role than syntactic structure.Flow chart models of the mental process in choosing word order patterns in English Chinese SI are proposed on the basis of the findings of this empirical study. On-line interpreters make judgments on the positions and the constituents of the segments before they conclude whether the SL information profile has been decided. This is the basis for deciding on the TL word order patterns, and memory load induced by the segments is an influencing factor.With variables strictly controlled, this dissertation investigates the patterns of TL word order in segments of noun phrases containing modifiers and the underlying reasons. The evidence-based conclusion sheds light on the reason, i.e. the definitiveness of the SL information profile, and the influencing factor, i.e. memory load induced by the segments before it proposes flow chart models of interpreters'mental process. This study may offer information to relevant research and interpreter training.